!The effect of M&As, joint ventures and strategic alliances on parent firm’s performance

!

The effect of M&As, joint ventures and

strategic alliances on parent firm’s performance

Evidence from subsidiary and parent level data in the agricultural industry

University of Groningen

Faculty of Economics and Business

MSc. Strategic Innovation Management

Name:

A.J.M. Zwanenberg

Student number: 2929066

Email address:

a.j.m.zwanenberg@student.rug.nl

Supervisor:

dr. K.J. McCarthy

Co-assessor:

dr. W.W.M.E. Schoenmakers

Date:

June 20, 2017

Abstract

Multinational corporations (MNCs) are constantly striving to obtain internal and external knowledge to exploit their position in the market. In the last decade, the role of the subsidiary has increased, and it is found that the subsidiary’s resources and capabilities can be leveraged within the MNC network. This paper combines three types of expansion strategies: mergers and acquisitions (M&As), joint ventures and strategic alliances to investigate the impact on parent firm performance. In particular, the paper examines whether an expansion strategy made through a subsidiary or parent firm has different effects on the parent firm performance. I employed different proxies for parent firm performance: cumulative abnormal returns, patents, return on assets and return on equity. Using an original dataset of the six largest agriculture firms for the period 2001-2012, I found evidence that there is no significant impact of M&As, joint ventures and strategic alliances (by subsidiaries and parent firms) on the parent’s patents level and cumulative abnormal returns. Furthermore, it is found that M&As by parent firms are negatively related to parent firm performance. Similarly, strategic alliances and joint ventures by subsidiaries negatively affect the parent firm performance. In contrast, it is empirically found that M&As by subsidiaries and strategic alliances by the parent firm both lead to higher firm performance, in terms of return on assets and return on equity. In addition, the findings reveal that cross-industry diversity negatively moderates expansion strategies on ROA and ROE. Cross-border diversity in M&As and strategic alliances, on the other hand, have proven to hamper the parent firm’s performance, whereas international joint ventures improve the firm performance.

Table(of(Contents(

1.( Introduction

5

2. Theoretical background

8

1. Motives for firm’s expansion strategies 8

2. Expansion strategies 9

2.1 M&As 10

2.2 Joint ventures and strategic alliances 11

3. Hypothesis development

12

1. Relevance of subsidiaries 12

1.1 Subsidiary-parent performance 13

2. Expansion strategies and the subsidiary-parent relationship 14

2.1 M&As 14 2.2 Strategic alliances 15 2.3 Joint ventures 16 3. Diversity 17 3.1 Cross-border diversity 17 3.2 Cross-industry diversity 19 4. Conceptual framework 20

4. Research design and methodology

21

1. Empirical setting 21

2. Data collection and sample selection 22

3. Variable specification 22

3.1 Dependent variable 23

3.1.1 Market-based measure 23

3.1.2 Accounting-performance measure 25

3.1.3 Qualitative performance measure 25

3.2 Independent variables 25 3.3 Control variables 26 3.4 Moderator variables 26 3.5 Empirical models 26 3.6 Verification of variables 27 3.7 Descriptive statistics 28

3.7.1 Qualitative and accounting-performance study 28

3.7.2 Market-based measure 29

5. Empirical results

30

1. Regression results of expansion strategies on firm performance 30

1.1 Market-based performance study 30

1.2 Accounting-performance study 31

1.3 Qualitative performance study 32

2. The impact of national and industrial diversity 33

7. Concluding remarks

38

9. References

40

10. Appendices

49

b) Correlation matrix (longitudinal study) 51

c) Correlation matrix (event study) 52

d) Hausman test 53

3. a) Regression results (PATENTS) 54

b) Regression results (ROA) 55

c) Regression results (ROE) 56

4. a) Regression results moderating effect (CAR) 57

b) Regression results moderating effect (PATENTS) 58

c) Regression results moderating effect (ROA) 59

d) Regression results moderating effect (ROE) 60

List of tables

Table 1: Sample selection 22

Table 2: Overview of variables per study 23

Table 3: Descriptive statistics 28

Table 4: Descriptive abnormal returns 29

Table 5: Descriptive statistics 29

Table 6: Regression coefficients (CAR) 30

Table 7: Regression coefficients (ROA and ROE) 31

Table 8: Regression coefficients (PAT) 32

Table 9: Regression coefficients moderating effect 34

Table 10: Summary of regression coefficient effects on dependent variable 35

Table 11: Moderating effects 37

List of graphs

Figure 1: A continuum of cooperative arrangements 10

Figure 2: Knowledge flows 13

Figure 3: Conceptual framework 20

Introduction(

The strategic decision-making process has shifted its attention since the 1990s. Since that time, scholars and managers were convinced that knowledge derives from both the internal and external environment, and this shift has proved to improve the firm’s global competitiveness (Baglieri et al, 2010). Especially multinational companies (MNC) carry out several types of strategic partnerships to benefit from economies of scale, taxation, market entry or complementary resources (Penrose, 1959; Kemal, 2011; Mulwa, 2015). A new development in this strategic process is the active role of subsidiaries in the global value chain since the 2000s. These, often globally dispersed, subsidiaries can obtain and share resources with competitors or distributors (horizontal collaboration) or upstream/downstream partners in the supply chain (vertical collaboration) (Guimon, 2013). Thereby, subsidiaries’ competences within the MNC network can be leveraged in intra-firm collaboration (Baglieri et al, 2010).

All existing literature about subsidiaries solely examines the organizational and managerial issues (investment motives, principal-agent and ownership topics), but neglects the financial impact (Hansen & Gwozdz, 2013). Although previous studies incorporate the rationale that MNCs acts as loosely coupled networks with autonomous subsidiaries that have their own business objectives and no interdependent activities with the parent company (Baglieri et al, 2010), recently it has been supported that subsidiaries are embedded in the host country systems via partnerships (e.g. alliances, joint ventures or other agreements) (Dunning, 1996; Harrigan, 1988). In turn, the parent company’s objective is to dominate global technological flows, as MNCs want to gain access to “globally dispersed resources, knowledge and capabilities” to become innovative and maintain its competitiveness (Baglieri et al, 2010, p.200). Besides, existing literature on partnering strategies has mainly been restricted to financial and competitive performances of the MNC parent, analyzed by the industry factors (e.g. Five

Forces Theory) (Porter, 1980), firm and management specific factors (e.g. resource based view, agency cost theory, free cash flow theory, empirical building theory, hybris theory) (Barney,

1986; Barney, 1991; Peteraf & Barney, 2003; Mueller, 1989; Mulwa, 2015; Jensen, 1986; Amihud & Lev, 1981; Roll, 1986), and market factors (e.g. transaction cost theory,

internalization theory) (Coase, 1937; Williamson, 1975).

This shows that existing papers about (strategic) partnerships and expansion strategies have increased the understanding of why and with whom firms create partnerships, but it tends to frame that all partnerships are formed by the ultimate parent company (monolithic view) (Bos et al, 2015). In this way, previous studies remain silent about the impact of subsidiary partnerships on parent firm performance. This paper was inspired by the lack of knowledge about the role of the subsidiaries in the collaboration process, and this clearly suggests that more research is needed to broaden the understanding of the impact of different collaboration types of subsidiaries and parent firms on MNC parent performances. Although that MNCs are reluctant to publish subsidiary data for tax purposes (Christmann et al, 1999) and different inter-firm accounting measures are used (Anderson et al, 2002), this Master thesis wants to provide theoretical and empirical fundamentals to why subsidiaries enter different types of partnerships, and its effect on parent firm performance. A variety of literature and available firm-level data that covers different drivers of firm performance and partnering strategies, respectively mergers, strategic alliances and joint ventures, are used to narrow this literature gap.

global workforce (Ramamurthy & de Lavigne, 2017). In the last 40 years, the industry has changed profoundly. The small family-owned businesses are nowadays dominated by a small number of transnational corporations. Especially since the 1990s, mergers and acquisitions has led to the rise of the Big Six corporations within the agricultural sector - BASF, Bayer, The Dow

Chemical Company, DuPont, Monsanto and Syngenta. These six companies are capturing

significant market shares in the global agrochemical market (75%), commercial seed market (63%) and private sector research in seeds and pesticides (>75%) in the input market (ETC Group, 2015). This dominant strategy enables to control the sectoral directions and future research and innovation programs (Shand, 2012). Moreover, they could improve the “global technology transfer and close the productivity gaps between regions and countries”, but also raise concerns in public policies and market competition, economies of scale, high entry barriers and the potential loss of innovations (Fuglie et al, 2013).

This sector is characterized by strong growth drivers, such as the population, urbanization and the rise of the middle classes worldwide in combination with strong public support. This indicates that the world food production must increase with approximately 74% by 2050, and thus drives dramatic changes in the sector as a whole (Ramamurthy & de Lavigne, 2017). However, the sector also experiences challenges because of “climate change, rapid technological innovation, new demands for biofuels and access to information, manifesting through an increased volatility, complexity and scrutiny throughout the value chain” (Stirling et al, 2013, p.3). These challenges require new strategies and business models to create new opportunities, adding more dimensions to an already complex scenario. This forces firms to source for technologies and knowledge externally by collaborating and cooperating between and across different players in the value chain. Agricultural firms seem to consider options like mergers and acquisitions, joint ventures and strategic alliances, “irrespectively of sector, size and type of operation” (Ramamurthy & de Lavigne, 2017, p.3).

The relationship between partnering strategies and parent firm performance has been validated using a cross-sectional, based on event study, and longitudinal dataset of the Big Six firms and its subsidiaries worldwide in the period 2001 - 2012. Firm-specific information is gathered from corporations’ annual reports and Orbis (Bureau van Dijk), whereas data about expansion strategies are obtained from the electronic database Thomson Securities Data Corporation (SDC) Platinum. It is expected that the broad set of mergers, strategic alliances and joint ventures of both the subsidiaries and the parent firms affect the parent firm behavior. The parent firm performance is measured with four different variables in this study: cumulative abnormal returns, patents, return on assets and return on equity. In this way, I can observe both the parent firm’s performances based on market-based, accounting-based and qualitative measurements. Consequently, this study examines the moderating effect of diversity on parent firm performance by elaborating on cross-border and cross-industry partners. As firm performance is determined by the firm’s resource and capability endowment (e.g. Penrose, 1959), gathering diverse knowledge, resources, new market/industry entry and technological competences are useful. In this perspective, both international border) partners and unrelated (cross-industry) partners play an essential role and improves the learning process. However, it is expected that diverse partnerships are more difficult to control and coordinate at a certain point, what accordingly decreases the partnership performance (Sampson, 2007).

return on assets. Similarly, strategic alliances and joint ventures by subsidiaries negatively affect the parent firm’s return on equity. In contrast, it is empirically found that M&As by subsidiaries and strategic alliances by the parent firm both lead to higher firm performance, in terms of return on assets and return on equity. The effect of cross-border and cross-industry partnering on parent firm performance shows some conflicting results. The expansion strategies are positively moderated by cross-industry diversity for cumulative abnormal returns and patents. In contrast, the effect of expansion strategies on return on equity and return on assets is negatively moderated by heterogeneous partnerships. Third, international partners in M&As and strategic alliances have proven to hamper the parent firm’s performance, due to the differences in culture, business environment, rules and regulations. However, international joint ventures improve the firm performance (return on assets and return on equity), as it is likely to bring complementary resources and market entry. These findings serve as directions for managerial implications. Management should be aware of the different effects of expansion strategies and financial performances. Thereby, managers should assess the performance based on the international and heterogeneous character of the partner (or target-firm), since these factors affect the synergy, integration and knowledge transfer process within the MNC network. My findings have essential implications for the MNC and strategic partnership literature. First, this thesis incorporates several theoretical insights from both internal (firm and managerial specific) and external (market) motives for expansions. Second, his study validates whether subsidiary’s expansion strategies react differently on parent firm performance than parent firm’s expansions. In this respect, this thesis contributes to the literature by shifting its focus to the subsidiary-level and parent-level. Third, the effect of cross-border and cross-industry diversity is complemented to find a possible explanation for international and intra-industry partnerships on parent firm performance. Fourth, this thesis has analyzed market-based, accounting-based and qualitative performance measures to define the parent firm’s successes. This allows me to integrate arguments from different dimensions. However, the paper also acknowledges some limitations, that can serve as directions for future research. Future research could consider unique sets of control variables for the dependent variables, allowing a larger or different sample, or focusing on a different time span. Moreover, including different regions and countries to measure the interaction with expansion strategies could validate the effects of nationality. In addition, including the motives for the expansion and using different methods for determining partner heterogeneity could contribute to the literature.

This paper is structured as follows: the theoretical background provides an overview of the theoretical and empirical literature on (motives for firm’s) expansion strategies. Here after, the third section includes the hypothesis development and conceptual framework. The fourth part formulates the sample selection, methodology and variable specification. The fifth section interprets the empirical regressions based on event study and cross-sectional time series. The sixth section discusses the empirical outcomes. Finally, the concluding remarks, limitations and future research are discussed.

Theoretical(background(

In recent decades, the firms are constantly examining what expansion and collaboration strategies provide the best opportunities to exploit their business. The globalization process, driven by the lower transportation, communication and information costs, have led to the ability that firms consider new geographical locations and strategic business partners to gain more competitive advantage and market share (Baldwin & Evenett, 2014; Martinez-Noya et al, 2012). The topic is widely discussed in the international business and international economics literature. Especially the drivers of financial and (global) competitive performance of MNCs are examined, because scholars have emphasized the importance of MNCs to bring knowledge, skills, market entry and technology within the ever-expanding global market (Ozawa, 1992). However, the empirical literature on firm partnering shows inconclusive results. Some scholars argue that partnerships lead to higher MNC firm performance and competitive advantage (e.g. Ferrer et al, 2013; Selden & Colvin, 2013; Puranam & Singh, 1999; Baum et al, 2000; Stuart et al, 1999), while others claim that partnerships destroy the firm value (e.g. Moeller et al, 2005; Siegel & Simons, 2010) by often using insights from the agency cost of hybris theory. This theoretical background has the aim to provide the insights from existing research on motives for firm’s expansion, and the theoretical fundamentals of M&As, joint ventures and strategic alliances.

1.( Motives for firm’s expansion strategies

shareholder value, because managers and shareholders have different interested that lead to self-interest (Mueller, 1989). The more hierarchical governance modes can control agency problems, and thus are M&As or joint ventures preferred over strategic alliance since partners set up a new legal entity and share the equity (Kogut, 1988; Pisano, 1989). This is indirectly involved in the collaboration consequence that the firm’s size and reputation change after partnering, what leads to greater influences in the market space (market control theory) (Mulwa, 2015). Moreover, diverting cash flows increases the company’s size, power and earnings by diverting free cash flows (free cash flow theory), and avoids managers to use external capital (Mulwa, 2015; Jensen, 1986). Second, the empirical building theory impacts the deals, as managers are willing to maximize the size of the company by doing acquisitions to increase personal power (Amihud & Lev, 1981). Third, the Hubris theory argues that overconfident managers are incapable of maximizing shareholder value (Roll, 1986). Therefore, managers are not rational and the gains of the acquisitions cannot offset the costs. On the other hand, several scholars point external factors in their studies. According to Coase (1937) and Williamson (1975), firms should outweigh whether the internalization (make) or market exchange (buy) strategy is the most relevant (Das & Teng, 2000). Firms are considering expansion (e.g. M&As) to internalize goods and resources to reduce the transaction costs, what is in line with higher corporate risks (Williamson, 1981; Buckley & Casson, 1976). Internalization leads to the creation or exploitation of the firm’s core competences (Dunning, 1977; 1985; 2000). Only firms that experience barriers to M&As and expect that benefits will outweigh the costs, will form alliances (Buckley & Casson, 1988; Geringer, 1991). In this way, collaborating is more cost efficient. However, transactions are subject to asset specificity that can hamper expansion performance. Asset specificity can be defined as “investments that have a higher value within a specific transaction relationship than outside the relationship”, and thus sunk cost if the firm’s relationship is distorted or terminated (Klein, n.d. ,p.1). Moreover, companies should be aware of asset specificity in alliances, as this brings high coordination costs. Hierarchical governance modes are more preferred due to the incentive alignment (Pisano, 1989; Klein et al, 1978). The current literature incorporates these insights, but also highlights that companies continually create strategies to respond to the effects of the globalization: the speed and costs of technological advancements in an uncertain business environment (Hagedoorn & Schakenraad, 1992). Firms are motivated the expand to gain access local market knowledge, institutional environment and management (Zekiri, 2016). Moreover, the access to resources and capabilities in the market enhances the firm’s performance by means of market share, lower costs and risk, increase in the firm’s bargaining power and other synergic opportunities that are essential for the firm’s survival and success (Inkphen & Beamish, 1997; Contractor et al, 2003; Kogut & Zander, 1992; Drees, 2014).

2.( Expansion strategies

to performance varies, because of the different firm’s motives. For example, a M&A strategy is preferred if the firm wishes the absorb all of the target firm’s resources, while a strategic alliance is formed if the firm wishes to provide access to knowledge from its partner while maintaining their own control (Drees, 2014). Moreover, a joint venture has a distinct entity, established by firms that are interested in joint development or technology sharing. In the following, the theoretical fundamentals of mergers and acquisitions, strategic alliances and joint ventures will be elaborated.

Figure 1 – A continuum of cooperative arrangements (Lorange & Roos, 1991). 2.2.1( M&As

An event in which two companies consolidate to form a new company with a new management and shares, is called a merger. Unlike, an acquisition is when the acquirer firm takes over the target firm. Theorists and empiricists investigated a broad variety of M&A disciplines and highlight each a different stage of the process: pre-merger, during-merger or post-merger stage (integration) (e.g. Chatterjee, 2009). Mergers and acquisitions are an often-used strategy, driven by grow business confidence, consumer demand and economic reasons (Mulwa, 2015). The main aim to adopt such a strategy is to increase the market share in a competitive market, benefit from economies of scale or taxation (abroad), and increase the customer base (Kemal, 2011; Mulwa, 2015). Moreover, Sharma (n.d.) argues that M&A activities have become popular due to the emerging players that entered the market, giving access to new capabilities, resources and market entry. In addition, the support by shareholders, strong financials, cheap debt and available cash have facilitated the latest M&A activities (Sharma, n.d.). Therefore, the industry players are conscious that a fruitful implementation of M&As can lead to high financial performances and a sublime market position.

2.2.2( Joint ventures and strategic alliances

Strategic alliances are flexible agreements between two (or more) independent organizations to share resources, capabilities and knowledge on a continuing basis to chase mutual goals (Das & Teng, 1998). While the companies maintain its autonomy, the benefits and control over the performance of the tasks are shared (Yoshino & Rangan, 1995). A joint venture, on the other hand, is a more permanent agreement. Two (or more) companies combine their resources and create a separate business entity (Harrigan, 1984; McGahan et al, 2016). According to PwC’s 2016 Global CEO Survey, 49 percent of the global CEOs expected to form partnerships (McGahan et al, 2016), as firms expect to benefit by gaining access to the partner’s complementary resources and capabilities (Van Beers & Zand, 2014). Moreover, knowledge, risk and cost sharing, economies of scale, reduction of the competition and learn opportunities from partners are drivers for alliances (Eisenhardt & Schoonhoven, 1996; Hagendoorn, 1993). Therefore, managing multiple alliances formations with diverse partners has become popular and firms are eager to exploit the external capabilities (Hoffmann, 2005).

The research perspective changed throughout the years. In the past, the effect of dyadic-level alliances on firm performance was studied (e.g. Sampson, 2007). Nowadays, firms have become more embedded in organizational networks (De Vries, 2008). Scholars are aware of this trend and have often examined the impact of alliance portfolios, the network level and its level of firm performance (e.g. Lahiri & Narayanan, 2013; Faems et al, 2012). Thereby, the alliance outcome focused mainly on outcome as effect of formation conditions (e.g. Park & Ungson, 1997) or collaboration (e.g. Larsson et al, 1998). While the first empirical studies used financial indicators such as growth, costs savings and profitability (e.g. Geringer, 1990; Lecraw, 1983), the latest studies focus on parent learning and the parent firm’s strategies (Kogut, 1988; Contractor & Lorange, 1988).

Hypothesis(development

The discussion in the previous section has made it evident that MNCs use different types of expansion strategies (e.g. strategic alliances, joint ventures and M&As) to get access to new capabilities and resources, market entry and economies of scale. MNC performance are often measured as the parent’s performances over time (Al-Matari et al, 2014), by either financial performance measures (e.g. stock price or profits), or by non-financial performance measures (e.g. stakeholder satisfaction or compensation) (Bayraktar et al, 2016). However, many researchers have created measurement models, none can measure every aspect clearly (Snow & Hrebiniak, 1980). In addition, the role of subsidiaries has often been overlooked.

In the following, the relevance of subsidiaries will be outlined. Further, based on the theoretical background and empirical findings mentioned in the previous section, I define five hypotheses to investigate the effect of expansion strategies of both subsidiaries and parent firms on MNC parent performance.

1.( Relevance of subsidiaries

The strategic decision-making process has shifted its attention since the 1990s. Since that time, scholars and managers were convinced that knowledge derives from both the internal and external environment. The role of subsidiaries has become influential in gaining access to technological capabilities and resources in the local market via either inter-firm collaboration (leveraging available competences within the MNC) or intra-firm collaboration (leveraging

available competences with competitors (horizontal) or upstream/downstream players in the supply chain (vertical)) (Hansen & Gwozdz, 2013).

environment and know-how to succeed the internal and external coordination. Therefore, lots of MNCs have implemented interdependent global teams to gain access to “globally dispersed resources, knowledge and capabilities to make R&D more innovative as well as more efficient” (Baglieri et al, 2010, p.200). This requires interdependent business units, mostly centralized by the parent firm (Birkinshaw & Fey, 2000). However, a high level of absorptive capacity and social capital is needed in subsidiaries to have effective knowledge transfers from peer subsidiaries and the parent firm (Chiang, 2007).

Figure 2 – Knowledge flows (Mudambi & Navarra, 2004)

1.1 Subsidiary-parent performance

subsidiaries (different beliefs, values and culture) (Schneider,1992; Daft & Weick, 1984). In another study, Hansen & Gwozdz (2013) emphasize the positive relationship between subsidiary and parent performance by elaborating on the environmental determinism to explain that parent and subsidiary performance is negatively affected by formal and informal institutions (institutional theory) and the degree of competition and bargaining power in the industry (Porter, 1980). Finally, the study of Gugler et al (2013) mainly examined the institutional determinants of domestic and foreign subsidiaries’ performance. They state that subsidiary and parent firm performance are positively related, as only the most productive firms have foreign direct investments and exploit the ownership, location, and internalization advantages (OLI paradigm) (Helpman et al, 2004; Dunning, 1977; 1985; 2000).

2.( Expansion strategies and the subsidiary-parent relationship 2.1 M&As

M&As are undertaken to maximize firm value, what can be achieved via several ways. First, M&As increase the firm’s growth (Sirower & O’Byrne, 1998). Second, operational synergies (e.g. cost reduction, market power) or financial synergies (e.g. lower cost of capital, tax benefits) with the target company can be realized (Gaughan, 2011). The literature points out that there is consensus between researchers about value creation for target firms in M&As (e.g. Schwert, 1996). However, scholars disagree about the financial outcomes for acquiring firms. For example, studies using market-based performance measurements (e.g. shareholder returns) are often negatively associated with M&A announcements due to the risks and inflexibility (Drees, 2014). Thus, the acquiring firm could report negative returns (e.g. Franks et al, 1991) or positive returns (Lang et al, 1989). The theories behind these results cover that acquisitions have the objective to maximize the growth in sales, assets or control that leads to a rise of shareholder value of the target firms, while those of the bidding firms would lose (non-value

maximizing behavior). In addition, the value maximizing motivation argues that shareholders

expect a positive economic gain from the acquisition, leading to positive abnormal returns for the target firm and at least normal returns for the acquiring firm (up to 7%) (e.g. Halpern, 1983; Beitel et al, 2004; Babanazarov, 2012; Wu, 2016). Moreover, Fuller et al (2002) found that company that buys from private firms or subsidiaries gain value. However, not all companies are able to capture these financial gains, as it depends on how the MNCs deal with risks and uncertainties in the post-transaction integration stage. Therefore, firms should first screen and evaluate the acquisition targets, and assess whether the target firm is qualified. According to Sharma (n.d.), companies should investigate whether “(1) there is a (strategic) fit between the acquiring and target company, (2) the target company is (organizational) attractive (e.g. market share, technology, growth potential) and (3) the feasibility of acquiring the selected target” (p. 5). If the MNC is able to minimize the risks, the consolidation can result in innovative products by combining R&D capabilities. In this way, a long-term sustainable profit growth is realizable (Wu, 2016). Especially large corporations are experienced with M&A processes, facilitating the knowledge transfer and integration between the acquired and target firm (Back & Krogh, 2002). However, Kilian & Schindler (2013) analyzed M&As from another perspective and conclude that “M&As can be seen as a market control mechanism because companies where resources are not used in the most efficient way are likely to be acquired” (p.5). This will lead to short-term shareholder wealth maximization, what will be solved by the market and management (Masulis et al, 2007). Consequently, the management will decrease the probability of take overs that lead to negative abnormal returns (Jensen & Ruback, 1983).

However, the extent to which parent firms benefit from the acquired resources and capabilities through their subsidiary depends on two main factors. First, a M&A strategy is a hierarchical form of collaboration, and is preferred if the firm wishes to absorb all the target firm’s resources (Drees, 2014). M&As avoid high transaction costs, including the transfer of tacit knowledge (Birkinshaw et al, 2000). As result, this facilitates the learning process and linkage of new knowledge to prior knowledge bases, resources and capabilities (Vaara et al, 2012). The second factor includes the degree of knowledge transfer to the parent. The parent firm should learn and coordinate the M&A process as the acquiring and target firm differ in terms of technology, knowledge and resources (Cloodt et al, 2006). As subsidiaries have a high degree of ownership and goal congruence, subsidiaries are generally more integrated in the MNC network with good communication and management systems. In turn, agency problems in the subsidiary-parent relationship will be low, whereas the inter-firm knowledge transfer is high. In this way, parent firms will have a higher performance if subsidiaries undertake M&As. This is in line with Gupta & Govindarajan (2000), who found that “the value of knowledge stock, the motivation to share knowledge, and the richness of transfer channels are positively related to the outflows of knowledge from a subsidiary whereas the capacity to absorb the incoming knowledge, the motivation to acquire knowledge, and the richness of transfer channels are positively associated with the inflows of knowledge into a subsidiary” (Huang et al, 2012, p. 45).

Hypothesis 1b: Subsidiary firm’s M&As are positively related to parent firm performance.

2.2 Strategic alliances

A strategic alliance is a two-way process, indicating that the partners need to collaborate and share their reciprocal inputs to become valuable (Ahuja, 2000; Das & Teng, 2000). This provides opportunities to complement each other’s knowledge and resources, capabilities, gain a foothold in an unknown market or industry, and share costs and risks (Volberda, 1996; Martin & Stiefelmeyer, 2001), while maintaining its own autonomy. This enables companies to focus on their core competences and outsource others. In turn, organizations create competitive advantage (Volbera, 1996).

There are also other reasons that firms participate in strategic alliances. As example, firms have considered strategic alliances as response to corporate downsizing in the 1990s (Kaus, 1999). Moreover, firms that experience strong competition in the market pushes them to ‘win the battles over competitors’. In response, they enter in multiple alliances and hope that some will lead to innovation. However, as the number of strategic alliances increases, the quality of screening and monitoring diminishes and the firm becomes exposed to mismatches or partner’s opportunistic behavior (Deeds & Hill, 1996).

Besides, firms have not only entered strategic alliances with firms operating in other markets, but firms also collaborate with each other (competitors) to reinforce their market power (Shand, 2012). Although this type of collaboration provides some space for opportunism, strategic alliances have an informal and flexible character that may lead to differentiated products that are produced because of the lower risks and costs in the alliance. Perry et al (2004) and Tebrani (2003) found that there is always a positive relationship between entering strategic alliances and firm performance.

As subsidiaries are embedded in the MNC network, it is expected that the subsidiaries are influenced by these relationship with the MNC firm. Besides the fact that subsidiaries should have motivations to enter strategic alliances (e.g. access to specialized or complementary resources that cannot be developed internally) and the need to find right partners, the subsidiary-parent relationship in the MNC network should be considered (Bos et al, 2015). This makes it possible that alliances can conflict with the interests of other subsidiaries or headquarter within the network, what dampens the parent firm performance (Parise & Casher, 2003; Bos et al, 2015). Moreover, strategic alliances are a hybrid form of collaboration, that may cause opportunistic behavior and lower partner commitment (Tsang, 2000). Accordingly, firms should avoid misappropriation (Pisano, 1990). To prevent unintended knowledge leakages, the parent firm should coordinate and control the knowledge flows (Bos et al, 2015). However, not all subsidiaries can be fully controlled and managed by the parent. Some parent MNCs are not able to perform these tasks, or the subsidiary (management) does not allow these headquarters’ supporting actions. This all creates imposed risks for subsidiaries entering strategic alliances, as there is a considerable chance of moral hazard problems, driven by the lack of partner’s willingness to share their knowledge and high opportunity costs (Sampson, 2007).

Hypothesis 2b: Subsidiary firm’s strategic alliances are negatively related to parent firm performance

2.3 Joint ventures

Pooling resources and capabilities under a new entity (joint venture), is a type of collaboration closer to a firm compared to strategic alliances. As a joint venture has a new formal joint management and assigned employees working in the separate entity, it facilitates transfer of complex and diverse knowledge (Sampson, 2007). In addition, MNCs move experts and technical employees of the parent firm to the joint venture. In turn, the firm can develop its own and develop its own skills, leading to competitive advantage (Sampson, 2007). In the same vein, Schaan (1983) argues that parent firm’s control on the joint venture lead to higher financial performance.

The joint venture literature is predominantly convinced of the positive effect between joint ventures and firm performance. Reynolds (2012) argues that joint ventures “are often transitory, intended to meet short-term goals of improving efficiency and market access” (p. 6). Especially innovative performance profits are higher relative to strategic alliances (Sampson, 2007). Besides the fact that joint ventures can increase the power over other firms (Pfeffer & Salancik, 1978), a variety of studies proved this outcome via different measurement methods. The success of joint ventures related to stock market valuation is determined (e.g. Das et al, 1998). The stock price performance is seen as a measurement to determine the short-term success of joint ventures (Das et al, 1998). Alternatively, Barney & Mackey (2005) state that joint ventures can lead to value creation and competitive advantage if the resources are not easy to copy.

Hypothesis 3a: Parent firm’s joint ventures are positively related to parent firm performance

subsidiary firms. Although there is less chance of partner’s opportunistic behavior, there is still some risks to create conflicting priorities with firms in the network or chance of opportunistic behavior. More specifically, there is risks of lower central coordination (Doz & Prahalad, 1984).

Hypothesis 3b: Subsidiary firm’s joint ventures are negatively related to parent firm performance

3.( Diversity

The diversity in M&As, joint ventures and strategic alliances is measured by the different firm’s characteristics among the partners. Diversity gives opportunities to obtain knowledge from diverse partners, but makes it hard to recognize knowledge and brings management difficulties (Jiang et al, 2010). Research argues that the characteristics of partners are more important than the number of collaborations (Hagedoorn & Schakenraad, 1994). The literature distinguishes several types of diversity, such as industry diversity and national diversity. Both concepts will be explained in the following.

3.1 Cross-border diversity

Partnering strategies involve companies from different nations. Cross-border diversity can have several advantages, such as market entry (Glaister & Buckley, 1996) and complementary resources (Tung, 1993). Unless that it is expected that cross-border diversity does not matter anymore due to the dramatic changes in the communication technology (Kuo & Fang, 2009), a large amount of literature shows conflicting results. Some studies are convinced about the positive relationship between cross-border strategies and firm performance. As example, the study of Charkabarti et al (2009) shows that the long-term financial performance is significantly positive for M&As, as mergers may create synergies and organizational capabilities that are necessary to operate in the overseas market. However, other authors point the negative relationship between cross-border diversity and performance, because there are also potential motives for conflicts due to language, cultural or business differences.

According to Ghemawat (2001), cross-border diversity does not only include cultural differences in the expansion strategy, but also differences in administration, geography and economy (CAGE-framework). This all will influence the integration, knowledge transfer and complexity of the partnering. MNEs have to overcome this diversity, what is associated with some costs. These costs are so called ‘Cost of Doing Business Abroad’ (CBDA) or the liability of foreignness (LOF) (e.g. Eden and Miller, 2004; Beugelsedijk et al, 2013). These terms are derived as the total additional costs for dealing with new rules, regulations and new cultures (Beugelsdijk et al; 2013), or social costs of entering and acceptance in a new market (Zaheer, 2002). Most macro-economic differences, such as differences in policies, economic systems and industry standards, cause conflicts in the societal, national, corporate, strategic and management culture (Parkhe, 1991; Jiang et al, 2010). As example, the different regulations and laws in intellectual property rights, political instability or different norms and routines (Kogut & Sing, 1988), could lead to conflicts and difficulties to obtain tacit knowledge (Qi et al, 2015). Consequently, this negatively affects the integration process (Jemison & Sitkin, 1986; Weber et al, 1996), increases the expansion costs (Parkhe, 1991) and could even further affect the international expansion of MNCs (Hutzschenreuter et al, 2010). Moreover, language and communication barriers lead to interaction difficulties and higher coordination costs (Tung, 1993).

revealed some different effect of international partnerships for the three expansion strategies in this study: M&As, joint ventures and strategic alliances.

Motives for cross-border M&As include the access to resources (e.g. patented technologies, managerial skills or marketing) of an existing foreign target firm (Errunza & Senbet, 1981; Shimizu et al, 2004). In addition, it can be a strategy to enter a new market and overcome the government rules and regulates. However, based on the transaction cost theory and resource based view, it is very difficult to transfer these intangible resources (Lin, 2015). In addition, Lin (2015) found evidence that the integration of strategic assets is negatively related to firm performance. Thereby, the paper of Grimpe & Hussinger (2008a) found that foreign firms are generally larger, and are thus linked to higher uncertainty and risks (Bertrand & Zuniga, 2006). Besides, the monitoring costs, there are also costs related to cultural and institutional difference and transmitting tacit resources and knowledge. In this way, firms want to execute M&As if the expected returns from cross-border transactions are higher than the domestic M&A deals. However, Lin (2015) found that most firms are not able to operate internationally in cross-border M&As due to organizational, technological and political differences. This makes it hard to adapt to the new local environment, and learn from both the new market and its target firm (Zaheer, 1995). Similar motives have been found for joint venture or strategic alliance strategies. These international strategic partnerships also give access to new technologies, markets or resources (Contractor & Lorange, 1988). These motives do not only cover business goals (e.g. profitability), but also qualitative goals (e.g. learning) (Kogut, 1988). Although joint ventures offer more commitment compared to strategic alliances, international joint ventures can better face uncertainty in foreign markets. Nevertheless, joint ventures enforce close ties with the foreign partner, creating more opportunities to exploit the new market or its resources. However, managerial difficulties could arise in the international partnership, that could hamper the mutual goals (Parkhe, 1993). In combination with the high uncertainty and risk in the foreign market, this requires control to utilize and implement the firm’s resources and capability efficiently (Lorange et al, 1986). Inefficient managerial control is the main cause of joint venture failures, what exposes risk of opportunism. This can lead to knowledge leakage to outsiders or competitive loss (Geringer & Hebert, 1989).

Therefore, it is expected that in cross-border M&As, joint ventures and strategic alliances, the MNC network must deal with liability of foreignness, because there exists a high probability of information asymmetry, management uncertainties and risk management costs in these partnerships (Qi et al, 2015). In this case, the MNC network constantly deals with new rules, regulations and new cultures due to the unfamiliarity will all firms involved in the operations. Thereby, the control and confidence of investors decreases (Guerin, 2003). To overcome these costs, firms should understand the impact of distance on the firm’s strategy. In addition, Cezar & Escobar (2015) found that firms should gain experiences to overcome liabilities of foreignness. Moreover, they found that developed economies are better in dealing with these cross-border matters, as they adapt more easily.

Hypothesis 4a: The effect M&As, joint ventures and strategic alliances on parent firm performance is negatively moderated by cross-border diversity.

!

strategic goals, behavior and intentions to the subsidiary firm (Tsang, 2000). On top of that, a loss of parent firm control in this process leads to opportunistic behavior of the subsidiary.

Hypothesis 4b: The effect of subsidiary firm’s expansions on parent firm performance is stronger negatively moderated by cross-border diversity than parent firm’s expansions.

3.2( Cross-industry diversity

Firms are not only interested in expansion strategies with cross-border partners, but they are also looking for partners from other industries. Partners (often competitors) from the same industry have an overlap in activities, causing imitation and absorptive capacity (Cohen & Levinthal, 1990). Thereby, similar partners are less likely to adopt novel innovations (Luo & Deng, 2009). The effect of partner heterogeneity and firm performance are not conclusive. Most authors emphasize the positive relationship (e.g. Jiang et al, 2010; De Man & Duysters, 2005; Deeds & Hill, 1996). Partners from other industries possess new specialized knowledge and (complementary) resources, and the partners will be more likely to share their information (Stasser & Titus, 1987; Powell et al, 1996). The access to this diverse resource pool will give information advantages by facing a variety of (new) technologies (Ahuja, 2000). In addition, it facilitates the firm’s entry into new markets (Jiang et al, 2010) and firms dealing with heterogeneous partners learn how to deal with business and technological uncertainty (Hoffman, 2007). In turn, firms engaged in heterogeneous partnerships experience higher growth rates, higher rates of product development and are more profitable (Gulati & Singh, 1998; Deeds & Hill, 1996). Others argue that cross-industry diversity hampers the firm performance, because different routines, management practices and a lack of overlap make it difficult to absorb and assimilate the new knowledge (e.g. Vasadueva & Anand, 2011). Consequently, there exist a lack of synergy and resource ‘fit’ that leads to lower firm performance (Goerzen & Beamish, 2005). According to Goerzen & Beamish (2005), there is a U-shape relationship between cross-industry diversity and firm performance. First, expansion strategies with industry-diverse partners hamper the value creation. Here after, as firms are more experienced with these relationships, companies will reach a minimum threshold of diversity effectiveness and start learning (Jiang et al, 2010). In another study, Noseleit & De Faria (2013) found that firms with different industrial backgrounds are a source for new, specialized knowledge, whereas unrelated partners cannot benefit from this due to the gap between the knowledge base between the partnered companies. Therefore, partners from a related industry have a certain overlap that is essential to create new knowledge, recognize, transfer and assimilate it. Consistently, Sampson (2007) found an inverted U-shaped relationship between partner’s industrial diversity and firm performance. Homogeneous partners are not able to create novel innovations, whereas heterogeneous partners encounter problems in knowledge transfer and resource allocation. Therefore, firms with a related partner are likely to create (financial) benefits from the partnering strategy.

Hypothesis 5a: The effect M&As, joint ventures and strategic alliances on parent firm performance is positively moderated by related cross-industry partners

Hypothesis 5b: The effect of subsidiary firm’s expansions on parent firm performance is weakened by cross-industry diversity compared to parent firm’s expansions.

4.( Conceptual framework

This conceptual framework gives an overview of all hypothesis formulated in the previous paragraphs. These hypotheses will be tested on firm-level by two measurement techniques: accounting measures and stock market performance measures.

Figure 3 - Conceptual framework

Research(design(and(methodology(

(

The aim of this study is to analyze the effect of M&As, strategic alliances and joint ventures, made through subsidiaries and parent firms, on parent firm’s performance. For this research, an event-study and cross-sectional time-series data, or so called longitudinal, approaches have been chosen. The event study assesses the impact of event announcements on the believe of whether the event will create of destroy value. Subsequently, this analysis shift the focus of attention on firm level measures.

1.( Empirical setting

I test my hypothesis using data from the agriculture industry. I use this industry for my analysis based on a few reasons. First, mentioned in the introduction section, the agriculture industry has changed dramatically in the last 40 years. The small family-owned businesses are nowadays dominated by a small number of transnational corporations. Second, the industry is characterized by declining productivity gains, strong urbanization and inefficient supply chains that are disrupting and putting pressure on the agricultural sector. Agribusiness firms need to increase their productivity and keep prices low (Ramamurthy & de Lavigne, 2017). Thereby, the complex and insufficient supply chains make it difficult for agribusiness companies to meet the conscious health diet of the (mainly developed) consumers, since “governments are tightening food standards and there is an increased demand for food that are certified as sustainable or demonstrate a level of environmental/social commitment” (Ramamurthy & de Lavigne, 2017, p. 6). In combination with the high price volatility, it requires companies to create new strategies and business models for more business opportunities. This drives firms to source for technologies and knowledge externally by collaborating and cooperating (such as M&As, joint ventures and strategic alliances) between and across different players in the value chain (Ramamurthy & de Lavigne, 2017) to secure the resources, improve their technology and meet the consumer demands. Oganesoff & Howard (2014) reported that the main participants in the agricultural market participate in partnerships and expansions to improve their future growth by obtaining complimentary products and innovative resources. Thereby, the latest expansion activities have been facilitated by the shareholder’s support, strong financials, cheap debt and available cash (Sharma, n.d.).

In order to identify companies in the agriculture industry, the dominant strategy of the six largest agricultural firms (BASF, Bayer, the Dow Chemical Company, E.I. Du Pont De Nemours

and Company, Monsanto and Syngenta) was noticed after reading plenty of (academic) articles

(e.g. Stirling et al, 2013; Ramamurthy & De Lavigne, 2017; ETC Group, 2015; Shand, 2012). These six companies are capturing significant market shares in the global agrochemical market (75%), commercial seed market (63%) and private sector research in seeds and pesticides (>75%) in the input market (ETC Group, 2015). Besides, the ability of the Big Six firms to influence both intra- and inter-industry firms is fascinating. This dominant position impacts the directions of innovation programs, and grabs attention from both the political and business world due to the limited market competition, economies of scale and high entry barriers (Fuglie et al, 2013). Although BASF, Bayer, The Dow Chemical Company and DuPont have existed for at least one century, Monsanto and Syngenta are relatively new firms. Monsanto is founded in 2000 after a spun off from Pharmacia & Upjohn, and is thus incorporated as a stand-alone subsidiary1

. Syngenta is formed in 2000 by the merger of Novartis Agribusiness and Zeneca

1_{Derived from Monsanto’s corporate website:}

Agrochemicals2

. Therefore, these six companies are a great mix of firms with relatively short and long experiences within and across the industry.

2.( Data collection and sample selection

I construct my sample using several sources of data. First, I obtained strategic alliances and joint venture data from the Thomson Securities Data Corporation (SDC) Platinum database. Consequently, I searched by the ultimate parent name for the Big Six firms from the years 2001 onwards. As this database has only data available until 2012, it resulted in a total of 215 strategic alliances and 108 joint ventures for this period. The samedatabase is also used to gather data about the Big Six firm’s mergers and acquisitions by using the parent’s CUSIP number. In total, 430 deals are considered for the period 2001-2012. Subsequently, a distinction was made whether the subsidiary or parent firm was involved in the alliance (for both the M&As, strategic alliance and joint venture data). The results are shown in table 1.

Type of alliance Sample size Merger and

acquisition

Subsidiary 229

Parent firm 201

Strategic alliance Subsidiary 116

Parent firm 99

Joint venture Subsidiary 53

Parent firm 55

Table 1 - Sample selection

Second, Datastream is used to collect daily returns of investments for the six companies and the local market indices (S&P 500, Dax 30 and the Swiss market). This data is incorporated in the event study, an empirical analysis that determines the statistically significant effect of an event announcement on the firm’s value. The underlying logic is to find abnormal returns that are attributable to an event by adjusting for the return that stems from the fluctuation of the firm as a whole. Third, the firm-level data was collected from the Orbis (Bureau van Dijk) database and the annual reports. Although some discrepancies in numbers exist, the data from the annual reports is leading, but missing data was, eventually, added from Orbis.

3.( Variable specification

This thesis executes two different studies: a cross-sectional (based on event study) and longitudinal study. The literature outlines a large set of variables that have been used in both types of studies to explain expansion strategies on firm performance. In the empirical studies, many variables have been used in varied combinations and different outcomes. In this thesis, both approaches use different dependent variables, as the cross-sectional study will show the effect of M&A, joint venture and strategic alliance announcements on the firm’s stock price. On the other hand, the longitudinal approach is used to identify whether these expansion strategies influence the firm’s patent, ROA and ROE level. These studies are complemented with the independent variables, that have been explained in the hypothesis development. In addition, control variables are added in this analysis to improve the validity of the research. On the next page, the variables used per study are shown in table 2, and described in appendix 1.

Event study & OLS (1) Longitidunal study (2) Dependent variables Stock price x Patents x ROA x ROE x Independent variables M&As (subsidiaries) x x M&As (parent) x x

Joint venture (subsidiaries) x x

Joint venture (parent) x x

Strategic alliance (subsidiaries) x x

Strategic alliance (parent) x x

Control variables

Size x x

Leverage x x

R&D intensity x x

Moderator variables Cross-border diversity x x

Cross-sector diversity x x

Table 2 - Overview of variables per study

3.1 Dependent variable

The dependent variable is a variable that is caused by other variables in the model. The dependent variable in my thesis is firm performance at parent firm level. This is a broad term and no single measurement model can fully account for all aspects of firm performance, nor a guideline is created (Snow & Hrebiniak, 1980). As example, accounting-based performance measures show several limitations. Accounting statements show past performance of a firm, and cannot be used to predict future performance. In addition, management teams could manipulate the financial results and the accounting statements are presented on a going-concern basis (Mihalovic, 2016, p.102). In contrast, stock prices will reflect all information in an efficient market and are not dependent on manipulation of the management team (Mihalovic, 2016). Finally, qualitative metrics are used to measure the performance of firms by considering technological process, innovation and outputs, but neglect the financial and numerical aspects. Therefore, I made a distinction between market-based performance (e.g. stock price), financial performances (e.g. ROA and ROE) and qualitative performance measures (e.g. patents). In this analysis, these three types are considered and thus permit me to compare the effects of expansion strategies across a wide range of performance indicators.

3.1.1. Market-based performance

First, the firm stock’s price, or the firm’s stock market value, is likely to adjust following an alliance announcement. I analyze whether there is an abnormal reaction associated with the ad hoc news concerning the expansion, what is measured by calculating the difference between the actual and expected return of a stock in a certain (event) period. As investors trade in response to (an announcement of) an event, and thus reflects the market’s perception, this analysis uses this short-term performance measure. Therefore, this measurement determines the market-based success of an expansion, and the event study method will be used for the first analysis. This paragraph describes the event study methodology based on MacKinlay (1997) and Campbell et al (1997).

Estimate the time line of an event

determining the abnormal returns. To calculate the abnormal returns of an event, a certain time interval, or so called event window, should be chosen. The day on which the partnership was announced, is labelled as day 0. Usually, a few days before and after the announcements are included in the event window (Hauwald, 2003), because most markets are imperfectly efficient. This indicates that it might be possible that markets have not received the information about the announcements during trading hours or that the information is revealed after the closing (MacKinlay, 1997). Since this study will analyze the short-term impact of announcements on the firm’s performance, I use a small event window that measures 5 consecutive days [-2,2] for the six agriculture firms by using the local market index (S&P 500, DAX 30 and the Swiss Market) (Zollo & Meier, 2008). The length of this event window is supported by Walker (2000) and McWilliams & Siegel (1997).

Normal returns are measured over a period that is different than the event window. This period is called the estimation window. In this study, I will use an estimation window of 30 days [-60, -30].

Figure 4 shows the timeline for the event study. At τ = 0, the event is announced. The event window is spread from T3 (2 days prior to announcement) to T4 (2 days after announcement). This indicates the 5-day event window. Moreover, the estimation window is from T1 (60 days prior to announcement) to T2 (30 days prior to announcement).

Figure 4 –Time line event study ! Computing and analyzing abnormal returns

The second step includes determining the abnormal returns, the change in the stock price of a firm’s stock price after the event. The abnormal returns are the actual return subtracted by the expected return if the event had not occurred. The following methodology is derived from Fama et al (1969). The individual abnormal returns can be calculated as:

AR

%=%R

%–%E%(R

)%%

Where: ARit = abnormal return for company i on day t, Rit = actual return for company i on day t and E(Rit) = expected return for company i on day t

Where: E(Rit) = expected return for company i on day t,

+

-

and

R

!

The average abnormal returns measure the abnormal returns for all companies, and is calculated as:

AAR

%=%

0

12

0

34/ it%

%

Where: AARt = the average abnormal returns on day t, N = sample size (6 companies in our

study) and ARit = abnormal return for company i on day t

To determine the abnormal returns over the entire event timeline (window), the cumulative abnormal returns (CARs) and cumulative average abnormal returns (CAARs) are defined in the following formulas.

CAR

=%

12

%

Where: CARit = the average abnormal returns for company i on day t, N = sample size (6

companies in our study) and ARit = abnormal return for company i on day t

CAAR

=%

0

712

6

34/ it%

Where: CAARit = the average abnormal returns for company i on day t, N = sample size (6

companies in our study) and CARit = abnormal return for company i on day t. This shows that

CAR is the sum of the abnormal returns (Duso et al, 2010). 3.1.2. Accounting-performance measures

The return on assets (ROA) is the mostly used accounting-based indicator to measure the company’s profitability relative to its assets. The higher the rate, the better the management efficiency in using assets to generate earnings of the management team (Khatab et al, 2011). Moreover, the return on equity (ROE) is another accounting-based indicator in empirical studies that analyzes firm performance. This variable shows the profits earned by the company relative to the shareholder equity. The ROE is often-used to compare the profitability of different companies within one industry (Helfert, 2001). In this way, a high return on equity indicates that the firm is likely to generate cash, since the managers invested well in the funds provided by investors (Hutten, n.d.).

3.1.3. Qualitative performance measure

Fourth, the level of PATENTS is used as indicator to compare the technological performance of companies (Bresman et al, 1999; Cantwell & Hodson, 1991), because patents reflect the temporal right of an invention. Thus, a patent shows the firm’s innovative output after the expansion. Although scholars often debate about the shortcomings of this variable (e.g. Griliches, 1990), it is generally accepted method.

3.2 Independent variables

study added joint ventures that are created with the parent firm (JV_PAR) or subsidiary (JV_SUB). Likewise, strategic alliances created by the parent firm (SA_PAR) or subsidiary (SA_SUB) are considered.

3.3(Control variables

To make the research more reliable, this study controls for several firm level characteristics that might impact the effect of expansion strategies (independent variables) on firm performance (dependent variables). As firm performance is highly dependent on scale effects, this study will firstly control for firm size (Lahiri & Narayanan, 2013; Vijayakumar & Tamizhselvan, 2010; Amato & Burson, 2007). On the one hand, larger firms have more resources (Li, 2013). On the other hand, larger firms are likely to have more experiences with expansions. These experienced firms generally have access to a larger resource pool compared to new firms, can use their know-how and are likely to have more experiences in the expansion process (Sorensen & Stuart, 2000). This all positively influence the performance (Kale et al, 2002). An appropriate way to determine firm size is using the number of employees (SIZE) at the end of the year. Moreover, R&D intensity (R&D_INT) is the ratio of total R&D expenditure to the total net sales of the firm. R&D expenditures are seen as investments in intangible assets, what is necessary to differentiate themselves from other firms in the industry (Chan et al, 2011; Ehie & Olibe, 2010). In addition, higher R&D expenditures positively affect the firm’s innovation level and firm’s performance (Cohen & Levinthal, 1990; Fama & French, 1993). R&D intensity as control variable is widely adopted in empirical studies (Beld, 2004). Finally, leverage (LEV) is measured as the total liabilities relative to the total assets. This variable has been frequently adopted as control variable to measure firm performance in empirical studies (e.g. Kyereoah-Coleman & Biepke, 2006). Scholars found evidence that leverage positively impacts the firm’s performance, as the firm’s own capital is relatively more expensive than borrowed funds (Fosu, 2013). Moreover, companies can benefit from interest deduction by using debt financing (Modigliani & Miller, 1963).

3.4 Moderator variables

Degree of relatedness is one of the most discussed variables in the alliance literature (e.g. Porter, 1980; Shimizu et al, 2004). Alliance diversity is a concept that includes national (cross-border) and sectorial (cross-sector) diversity in this study. The degree of industry relatedness (IND) was obtained by comparing SIC (standard industrial classification) codes of the partner(s) (acquiring and target firm in case of M&As). Each partnership is measured on a scale from 0 to 4 (Rumelt, 1974). The partnership is coded 0 if the partners have the same 4-digit SIC code (SIC0), 1 if they have the same 3-digit SIC code (SIC1), 2 if they have the same 2-digit SIC code (SIC2), 3 if they have the same 1-digit SIC code (SIC3), and 4 if the SIC codes are completely different (SIC4). In addition, the degree of national relatedness (NAT) is determined by using value 1 if at least one of the partners has an international background, whereas value 0 is used if the partners (or target and acquiring firm in case of M&As) have the same nationality.

3.5 Empirical models

The studies of this thesis consist of firm-level data that explains the relationship of expansion strategies on firm level behavior. The literature includes a large set of variables that have been used to explain firm performance. The previous sections showed the variables (dependent, independent, control and moderator) for this study. The empirical model for the study has the following base:

These key measurements are constructed, and lead to the following empirical specifications.

Y

<QNC_DCE9:+ %<RNC_NOP9:+ <STUVONWE8_9:+ %<XUCWTYUCZTW8_9: J<[=9: + %\9:

[

where Y consists of four dependent variables: CAR (cumulative abnormal returns), PAT, ROA and ROE. The subscripts stand for firm (i) and year of observation (t). The betas are estimated parameters, and X is a vector of firm-specific control variables (SIZE, RD_INT, LEV). Finally, ] is the error term.

3.6 Verification of variables

This study uses a panel data set, because it combines both cross-sectional and time-series data for the dependent variables patents, return on assets and return on equity. Although most economic and business studies use Pooled Ordinary Least Squares (POLS), better technique to analyze this type of data set are: fixed effect or random effect model. Where the fixed effect model can explore the impact of individuals’ units that are time invariant (Clarke et al, 2010), “the random effect models allow for the possibility that population parameters vary from study to study” (Song et al, 2008, p.11). The Hausman test to will be used to decide which technique is the most appropriate in this study, as the null hypothesis suggest that the random effect model is reliable against the alternative hypothesis (fixed effect model is reliable) (Adkins & Hill, 2011). Consequently, by using the Pooled Ordinary Least Square (POLS) regression, this study will examine the relationship between the independent variables and firm performance. The analytical models will show a hierarchical pattern, as the first model will be regressed using the control variables and dependent variable only. This basic model will be complemented with the independent and moderating variables simultaneously. Finally, the full effect of expansion strategies on firm performance will be captured in the final model.